Rapid Diagnostic Tests (RDTs) are a critical frontline tool in the fight against malaria, with Community Health Workers (CHWs) playing a central role in their use across underserved communities. Yet diagnostic and interpretation errors remain a real concern, leading to wrong diagnosis and unnecessary commodity wastage.
To address this challenge, the Ministries of Health in Kenya and Uganda are implementing a multi-partner febrile surveillance study. This initiative integrates an AI-powered Malaria RDT Reader into eCHIS workflows to enhance diagnostic precision, surveillance and treatment at the community level. The study has adopted a rigorous mixed-methods randomized controlled trial design — combining quantitative and qualitative approaches. A total of 1,880 Community Health Promoters (CHPs) in Kenya and 1,863 Village Health Teams (VHTs) in Uganda have been enrolled and grouped into blinded and non-blinded arms. For the blinded group, malaria result interpretation happens seamlessly in the background, with only supervisors having real-time access to the RDT test results.
During the March Round-up, the project team and the implementing partners demonstrated the seamless integration of AI into the eCHIS assessment workflows:
- Standard interpretation: A CHW is able to interpret the RDT results as usual.
- AI Verification, a CHW captures the photo of the test results which is validated by the AI Reader.
- The AI Reader has multimodal capability which provides real time interpretation for validation.
This study represents a powerful example of how AI, community health systems and government collaboration can come together to drive meaningful improvements in diagnostic accuracy and health outcomes for last-mile populations. We are excited to continue learning alongside the Kenya and Uganda teams and the broader CHT community! You watch the presentation highlights here.